A Report Generator Volume 2
Abstract
The software is listed according to the flow of information in the system. First, the main module, which loads in all the the sub-modules, is presented. Following this are he mechanisms which interface the linguistical representations with the underlying knowledge representation. Then text strategies are illustrated followed by software for semantic, relational and syntactic analysis. A significant investigation was performed on the development of a successful focus selection algorithm, which also presented. The generation module has a feature dictionary system together with unification routines for the comparison of syntactic features and generation of well found grammatical constituents. Finally, the realisation routines are presented (including linearisation, morphological synthesis and orthographic formating). The knowledge sources for testing are presented followed by some system output. All software is well documented and includes a title, purpose, copywrite, and often a linguistic theory or principle behind the code. Each function is described in English so that even hacker's who speak dialects other than LISP can understand. Keywords: Natural language generators, Great Britain. (KR)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1988
- Accession Number
- ADA197330
Entities
People
- Mark T. Maybury
Organizations
- Air Force Institute of Technology